Manually tuning a musical instrument can be a difficult and tedious process, usually requiring a considerable amount of time and skill. Although having an automatic tuning system is desirable for ease and convenience, as well as for accuracy, there is another important reason. Frequently, a musician will need to change the tuning of an instrument during a performance or an instrument will go out of tune during a performance. And, during this process, it may be necessary to compensate for a change in an instrument's characteristics. For example, during a performance with a guitar, a musician may install a capo between selections. A capo is a device for clamping all strings to a particular fret, thereby increasing the frequencies of all strings by a constant factor. Because of the time required, manually retuning an instrument during a performance is usually unacceptable. One common, although expensive and inconvenient, solution to this problem is to have properly tuned spare instruments available for such occasions. A much better solution is to have a system for automatically tuning an instrument within a length of time short enough to be unnoticed by an audience.
Many different types of automatic tuning systems have been devised. There are open-loop systems which drive a tuning actuator to a predetermined position for each desired frequency. These have the advantage of being able to change tuning, silently and therefore unnoticed, during a performance. However, they have the disadvantage of being only as accurate as the predicted relationship between the frequency of the tone produced by the instrument and the actuator position.
There are closed-loop systems which measure the frequency of the tone produced by the instrument, compare it to a desired value, and use the result of the comparison to control an actuator which tunes the instrument. This technique is accurate in that it directly controls the frequency of the instrument and is independent of other factors which affect frequency. However, it has the disadvantage that an audible tone must be produced while the instrument is being tuned; and that audible tone generally precludes tuning during a performance.
Some stringed instrument systems, because of interactions between strings, sequentially tune each string and then iterate to compensate for the interactions. Others tune selected strings, or all strings, simultaneously and then iterate. These techniques require producing a tone, taking a frequency measurement, estimating and executing an actuator movement, then taking a new frequency measurement and repeating the process until the frequency produced is sufficiently close to the desired frequency.
Other systems measure the tension of (actually, the force applied to) a string and compare the measured value with a desired value to produce an actuator control signal. Although the string tension method does not require a tone to be produced while tuning, it does require a known and stable relationship between string tension and frequency. Satisfying this relationship requirement is difficult because frequency also depends on string length and mass per unit length as well as other factors.
A typical stringed musical instrument has a semi-rigid structure which changes form slightly when string tensions in the instrument are adjusted during tuning. A change in form due to the adjustment of one string therefore affects the frequencies of the remaining strings. Temperature and humidity also affect the form, and the frequencies, of the instrument in more subtle ways.
A system which compensates for the effect of adjusting one string on the frequencies of the remaining strings, described in U.S. Pat. Nos. 4,803,908 and 4,909,126 to Skinn et al., which are incorporated by reference herein in their entirety, involves the use of a calibration function which relates the position of each actuator to the frequencies produced by all the instrument's strings. Creating the calibration function involves the measurement of frequencies at multiple positions of each actuator and, through regression techniques, relating the position of each actuator to not only the frequency of its own string but to the frequencies of the other strings as well. The use of regression techniques provides the advantage that a priori knowledge of the detailed characteristics of the instrument being tuned is not required. Also, the calibration function can be updated by recalibration as the instrument ages, or as environmental or other changes occur. Using a calibration function generated from the particular instrument being tuned permits open-loop, and therefore silent, tuning with accuracy comparable to that of closed-loop systems.
In all of the previously described open-loop systems, a calibration function relates a desired frequency to an actuator position. However, if such a system is calibrated without a capo and then a capo is installed, this relationship is destroyed and the system must be recalibrated for each position of the capo. It is therefore an object of this invention to provide for automatically tuning a stringed musical instrument after installing a capo without having to recalibrate the system.
In closed-loop systems wherein the measured frequency is compared to a desired frequency, installation of a capo shifts the measured frequency relative to the desired open string frequency and thereby skews the comparison. It is therefore a further object of this invention to provide for automatic closed-loop tuning of a string instrument after installing a capo.